Printer data monitoring and early warning system based on artificial intelligence

Abstract

The invention provides a printer data monitoring and early warning system based on artificial intelligence, which relates to the field of printer monitoring technology and artificial intelligence application technology, and comprises a data acquisition module, an edge calculation module, an AI core analysis module, an early warning linkage module and a data storage module, the data acquisition module is a multi-source data sensing unit and outputs a structured data stream after preprocessing; the edge computing module is deployed at a local edge computing node of the printer component; and the AI core analysis module is deployed in a cloud server. The detection and early warning system breaks through single parameter monitoring limitation, integrates hardware, environment and operation multi-source data, combines edge calculation and cloud AI collaborative analysis, realizes conversion from passive alarm to active prediction, improves monitoring precision under different working conditions, and solves the problems that a printer monitoring technology is single in monitoring dimension, only pays attention to the state of hardware consumables, and is poor in monitoring precision. And core components of the printer are neglected.

Description

Technical Field

[0001] This invention relates to the fields of printer monitoring technology and artificial intelligence application technology, and in particular to an artificial intelligence-based printer data monitoring and early warning system. Background Technology

[0002] As a core output device in office and production scenarios, the stability of printers directly affects work efficiency.

[0003] Existing printer monitoring technologies have a single monitoring dimension, focusing only on the status of hardware consumables and neglecting the operating parameters of core printer components such as printheads, rollers, and circuit boards, as well as the impact of environmental factors such as temperature, humidity, and dust concentration on the equipment. This leads to one-sided fault prediction, and the early warning accuracy of the monitoring technology is low. Fixed thresholds cannot adapt to the differences in printer operation under different usage frequencies and load intensities, which easily leads to false alarms or missed alarms. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an artificial intelligence-based printer data monitoring and early warning system to solve the problems mentioned in the background section.

[0005] To solve the above-mentioned technical problems, the present invention is achieved through the following technical solution: This invention is an artificial intelligence-based printer data monitoring and early warning system, specifically including: a data acquisition module, an edge computing module, an AI core analysis module, an early warning linkage module, and a data storage module. Each module achieves data interaction through the TCP / IP communication protocol. The data acquisition module is a multi-source data sensing unit, which includes: working parameter sensors, environmental parameter sensors and operation behavior acquisition units, and outputs a structured data stream after preprocessing. The edge computing module is deployed on the local edge computing node of the printer component and is responsible for data processing, data diversion, real-time threshold judgment and data compression. The AI ​​core analysis module is deployed on a cloud server and includes: a feature fusion submodule, a deep learning prediction submodule, and a dynamic threshold generation submodule. The early warning linkage module, based on the output of the AI ​​core analysis module, realizes hierarchical early warning and multi-channel linkage, and links the printer control system, operation and maintenance management platform and consumable management system to perform response operations. The data storage module adopts a hybrid architecture of edge storage and cloud storage, and stores real-time data and historical data in a hierarchical manner.

[0006] Furthermore, the operating parameter sensor of the data acquisition module is integrated inside the printer assembly and into the printer assembly's built-in API interface, and the operating parameter sensor collects operating data of the core components.

[0007] Furthermore, the environmental parameter sensors of the data acquisition module are deployed around the printer assembly, and the environmental parameter sensors collect ambient temperature and humidity, dust concentration and voltage stability; the operation behavior acquisition unit obtains user operation data through the control chip inside the printer assembly; The data acquisition module denoises, standardizes, and removes outliers from the acquired raw data to generate a structured data stream.

[0008] Furthermore, the edge computing module processes the acquired data, including: data splitting, real-time threshold judgment, and data compression; data splitting performs local calculations on high-frequency real-time data and uploads low-frequency statistical data to the cloud AI core analysis module; real-time threshold judgment uses a built-in lightweight rule engine, presets basic thresholds based on printer model, and quickly judges sudden extreme value data to trigger immediate warnings; data compression uses the LZ4 compression algorithm to compress the uploaded data.

[0009] Furthermore, the feature fusion submodule of the AI ​​core analysis module uses an attention mechanism to fuse multi-dimensional data features, assigning dynamic weights to hardware parameters, environmental data, and operational behavior features to generate a comprehensive feature vector.

[0010] Furthermore, the deep learning prediction submodule of the AI ​​core analysis module constructs a hybrid model based on LSTM and CNN, where CNN is used to extract spatial features from the data and LSTM is used to capture time series features; the hybrid model of LSTM and CNN is trained and optimized through a large number of printer component failure samples and common failures, with the input being a comprehensive feature vector and the output being the probability of failure and the failure type in the next hour. The dynamic threshold generation submodule of the AI ​​core analysis module dynamically adjusts the warning thresholds of each parameter based on the printer's historical operating data and real-time operating conditions using a random forest algorithm.

[0011] Furthermore, the warning linkage module classifies warnings into three levels based on the probability of fault occurrence and the degree of impact. The warning levels are as follows: Level 1, only indicated by the indicator light on the operation panel of the printer component, and optimization suggestions are pushed simultaneously; Level 2, potential fault, notifying the equipment administrator via SMS and WeChat, with a fault prediction report attached; Level 3, immediately triggering an audible and visual alarm, automatically pushing dispatch information to the operation and maintenance management platform, and suspending non-emergency printing tasks. The early warning linkage module is linked with the printer control system. When a level 1 early warning occurs, the printing parameters are automatically adjusted. When a level 3 early warning occurs, a safety protection operation is performed. The linkage processing is also linked with the consumables management system. When a consumables depletion-related fault is predicted, the consumables procurement process is automatically triggered.

[0012] Furthermore, the edge storage of the data storage module stores high-frequency real-time data for the past 7 days locally using SSD storage to ensure fast data read and write speeds; the cloud storage of the data storage module uses a distributed database to store historical data, providing data support for model iteration and operation and maintenance optimization.

[0013] Furthermore, an operation panel is installed on the top of the printer assembly, and a storage side groove is provided on the side of the printer assembly; a guide rail is installed inside the storage side groove; an installation groove is provided on the side end of the guide rail, and a positioning slider is slidably installed inside the guide rail.

[0014] Furthermore, an environmental parameter sensor is installed at the outer end of the positioning slider; the environmental parameter sensor is located inside the storage side groove; and an operating parameter sensor is installed inside the printer assembly.

[0015] This invention provides an artificial intelligence-based printer data monitoring and early warning system, which has the following beneficial effects: When in use, this invention's detection and early warning system breaks through the limitations of single-parameter monitoring, integrates multi-source data from hardware, environment, and operation, and combines edge computing and cloud AI collaborative analysis to achieve a transformation from passive alarm to proactive prediction. The dynamic threshold generation mechanism solves the problem of poor adaptability of fixed thresholds and improves the monitoring accuracy under different working conditions.

[0016] By employing a hybrid LSTM and CNN model to mine the spatiotemporal features of the data and combining an attention mechanism to achieve dynamic feature weighting, the fault prediction accuracy reaches over 92%, which is 60% higher than the traditional threshold method. An integrated process of early warning, linkage, and processing is constructed, reducing the fault response time from an average of 2 hours to less than 15 minutes.

[0017] By reducing interference from invalid warnings through tiered early warning systems, the work efficiency of equipment administrators is greatly improved, the security protection mechanism reduces the risk of fault escalation, printer maintenance costs are reduced by 35%, and historical data mining supports the optimization of operation and maintenance strategies, extending the service life of equipment. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings of the embodiments will be briefly described below.

[0019] The accompanying drawings described below are only related to some embodiments of the invention and are not intended to limit the invention.

[0020] In the attached diagram: Figure 1 The overall system flowchart of the present invention is shown; Figure 2 A flowchart illustrating the workflow of the edge computing module of the present invention is shown. Figure 3 This invention illustrates the workflow flowchart of the AI ​​core analysis module. Figure 4 A flowchart illustrating the workflow of the early warning linkage module of the present invention is shown. Figure 5 A three-dimensional structural diagram of the printer assembly of the present invention is shown; Figure 6 A three-dimensional structural schematic diagram of the environmental parameter sensor of the present invention is shown; Figure 7 A three-dimensional structural diagram of the positioning slider of the present invention is shown.

[0021] List of reference numerals 1. Data acquisition module; 2. Edge computing module; 3. AI core analysis module; 4. Early warning and linkage module; 5. Data storage module; 6. Printer assembly; 601. Storage side slot; 602. Guide rail; 603. Mounting slot; 604. Positioning slider; 605. Environmental parameter sensor; 606. Operating parameter sensor. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the described embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] Please refer to Figures 1 to 7 : Example 1: This invention proposes an artificial intelligence-based printer data monitoring and early warning system, comprising: a data acquisition module 1, an edge computing module 2, an AI core analysis module 3, an early warning linkage module 4, and a data storage module 5. Each module interacts with the other via a TCP / IP communication protocol. The data acquisition module 1 is a multi-source data sensing unit, including: a working parameter sensor 606, an environmental parameter sensor 605, and an operation behavior acquisition unit. After preprocessing, it outputs a structured data stream. The edge computing module 2 is deployed on the local edge computing node of the printer component 6, responsible for data processing, data distribution, and real-time thresholding. Judgment and data compression; AI core analysis module 3 is deployed on a cloud server, and includes: a feature fusion submodule, a deep learning prediction submodule, and a dynamic threshold generation submodule; the early warning linkage module 4, based on the output of AI core analysis module 3, realizes hierarchical early warning and multi-channel linkage, and links the printer control system, operation and maintenance management platform, and consumable management system to perform response operations; the data storage module 5 adopts a hybrid architecture of edge storage and cloud storage, and stores real-time data and historical data hierarchically; the working parameter sensor 606 of the data acquisition module 1 is integrated into the printer component 6 and is built into the printer component 6. The API interface allows the operating parameter sensor 606 to collect core component operating data, including printhead temperature, roller speed, circuit board operating current, paper feed pressure, ink tank pressure, and consumable balance, with a sampling frequency of 10Hz. The environmental parameter sensor 605 of the data acquisition module 1 is deployed around the printer assembly 6, collecting ambient temperature and humidity, dust concentration, and voltage stability data, with a sampling frequency of 1Hz. The operation behavior acquisition unit obtains user operation data through the control chip inside the printer assembly 6, including print job type, print parameters, continuous working duration, and abnormal operation records, with the sampling frequency synchronized with the print job. The data acquisition module 1 performs noise reduction, standardization, and outlier removal on the acquired raw data to generate a structured data stream. The edge computing module 2 processes the acquired data, including data splitting, real-time threshold judgment, and data compression. Data splitting performs local calculations on high-frequency real-time data and uploads low-frequency statistical data to the cloud AI core analysis module 3. The real-time threshold judgment uses a built-in lightweight rule engine, presets basic thresholds based on printer model, and quickly judges sudden extreme value data to trigger immediate warnings. Data compression uses the LZ4 compression algorithm to compress the uploaded data, reducing network transmission bandwidth usage, with a compression ratio of 4:1. The feature fusion submodule of AI Core Analysis Module 3 uses an attention mechanism to fuse multi-dimensional data features, assigning dynamic weights to hardware parameters, environmental data, and operational behavior features to generate a comprehensive feature vector. The deep learning prediction submodule of AI Core Analysis Module 3 constructs a hybrid model based on LSTM and CNN, where CNN is used to extract spatial features from the data, such as the correlation between parameters of different components, and LSTM is used to capture time-series features. This hybrid model is trained and optimized using a massive number of printer component 6 fault samples, covering more than 20 common faults such as paper jams, printhead blockages, and circuit faults. The input is a comprehensive feature vector, and the output is the probability of fault occurrence and fault type within the next hour. The dynamic threshold generation submodule of AI Core Analysis Module 3 dynamically adjusts the warning thresholds of various parameters based on the printer's historical operating data and real-time operating conditions using a random forest algorithm. For example, during continuous high-load operation, the printhead temperature warning threshold is adjusted from 60℃ to 55℃; for new printer component 6, the vibration warning threshold is lowered during the initial use to adapt to the characteristics of the equipment break-in period. The warning linkage module 4's warning classification divides warnings into three levels based on the probability of fault occurrence and the degree of impact. The alert levels are as follows: Level 1, with a warning probability of <30%, indicates a minor anomaly, indicated only by the indicator lights on the operation panel of printer component 6, and optimization suggestions are pushed simultaneously, such as reducing printing speed to reduce component load; Level 2, with a warning probability of 30% ≤ warning probability <70%, indicates a potential fault, and the equipment administrator is notified via SMS and WeChat, along with a fault prediction report; Level 3, with a warning probability of ≥70%, indicates an impending fault, immediately triggering an audible and visual alarm, automatically pushing dispatch information to the operation and maintenance management platform, and suspending non-emergency printing tasks; The alert linkage module 4 is linked with the printer control system, automatically adjusting printing parameters during Level 1 warnings and performing safety protection operations during Level 3 warnings; The linkage processing is also linked with the consumables management system, automatically triggering the consumables procurement process when a fault is predicted to be of consumable depletion type; The edge storage of the data storage module 5 stores high-frequency real-time data for the past 7 days locally, using SSD storage to ensure data read and write speeds; The cloud storage of the data storage module 5 uses a distributed database to store historical data, including raw data, analysis results, and fault records, with a data retention period of 3 years, supporting multi-dimensional queries by device number, time range, and fault type, providing data support for model iteration and operation and maintenance optimization.

[0024] In this embodiment of the invention, data such as printhead temperature, roller vibration, and ambient humidity are collected in real time using a working parameter sensor 606 and an environmental parameter sensor 605. The data acquisition module 1's built-in data preprocessing subunit uses a Kalman filter algorithm to denoise the collected data and generate a standardized data stream, which is then sent to the edge computing module 2. The edge computing module 2 performs local analysis on the collected vibration data and finds that the vibration acceleration reaches 6.2 m / s² at a certain moment, exceeding the basic threshold. It immediately alerts the printer panel with a red light and simultaneously compresses data such as ambient humidity and continuous printing time before uploading it to the AI ​​core analysis module 3 in the cloud. The feature fusion submodule of the AI ​​core analysis module 3 processes the uploaded data... Weights are assigned to generate a comprehensive feature vector. After the LSTM and CNN hybrid model is input into this vector, it outputs that the probability of paper jam failure in the roller within the next hour is 68%. The dynamic threshold generation submodule lowers the roller pressure warning threshold from 0.8MPa to 0.6MPa. The system then triggers the secondary warning of the warning linkage module 4 and pushes the warning information to the equipment administrator via WeChat. The operation and maintenance management platform automatically generates a dispatch task and marks the required tools. At the same time, the system sends an optimization instruction to the printer to reduce the printing speed from 35 pages / minute to 25 pages / minute. The warning data, processing results and subsequent operation and maintenance records are synchronously stored in the data storage module 5 as samples for model iteration to improve the accuracy of subsequent predictions.

[0025] In Example 2, based on Example 1, an operation panel is installed on the top of the printer assembly 6, and a storage side groove 601 is provided on the side of the printer assembly 6; a guide rail 602 is installed inside the storage side groove 601; a mounting groove 603 is provided on the side end of the guide rail 602, and a positioning slider 604 is slidably installed inside the guide rail 602; an environmental parameter sensor 605 is installed on the outer end of the positioning slider 604; the environmental parameter sensor 605 is located inside the storage side groove 601; a working parameter sensor 606 is installed inside the printer assembly 6. When detecting the parameters of the printer assembly 6, the working parameter sensor 606 inside the printer assembly 6 integrates a temperature sensor, a vibration sensor, and a current sensor. The printer includes a pressure sensor and connects to the printer's built-in API interface. The operating parameter sensor 606 collects operating data of the printer's core components, including printhead temperature, roller speed, circuit board operating current, paper feed pressure, ink tank pressure, and consumable balance. The environmental parameter sensor 605 drives the positioning slider 604 to slide inside the guide rail 602 through the mounting slot 603, so that the environmental parameter sensor 605 is located inside the storage side slot 601, which does not affect the neatness of the printer component 6 surface. The two environmental parameter sensors 605 are in appropriate positions, so that the environmental parameter sensors 605 can collect parameters such as ambient temperature, humidity, and dust concentration to obtain the environmental parameters around the printer, which is convenient for generating standardized data streams.

[0026] Example 3, based on Example 1, describes a laser printer in an enterprise office setting. It includes various types of operating parameter sensors 606, a PT100 temperature sensor near the printhead, a piezoelectric vibration sensor at the roller bearing, a Hall current sensor connected in series at the power interface, and various environmental parameter sensors 605 such as a DHT22 temperature and humidity sensor and a PM2.5 dust sensor on the side of the printer. A data acquisition terminal is connected via the printer's USB interface to standardize the data. The data is then transmitted to the printer via a network cable through an edge computing module 2 based on the ARM Cortex-A72 architecture. A basic threshold for this printer model is preset. An AI core analysis module 3 is deployed on an Alibaba Cloud server, using the TensorFlow framework to build an LSTM-CNN hybrid model. The training dataset includes one year's worth of operational data from 1000 printers of this model and 5000 fault samples. The data storage module 5 uses a combination of Alibaba Cloud HBase database and local SSD, with an automatic data synchronization mechanism configured. The early warning linkage module 4 is integrated with the enterprise WeChat API, the operation and maintenance management platform, and the consumables procurement system, with early warning information push rules and dispatch processes configured.

[0027] The working principle of this embodiment is as follows: Data on printer operating parameters and environmental parameters are collected by the operating parameter sensor 606 and the environmental parameter sensor 605. The data acquisition module 1 denoises the collected data to generate a standardized data stream, which is then sent to the edge computing module 2 for local analysis. If the data exceeds the basic threshold, a red light on the printer panel will immediately indicate this. At the same time, the environmental data and printer operating data are compressed and uploaded to the AI ​​core analysis module 3 in the cloud. The AI ​​core analysis module 3 assigns weights to the uploaded data to generate a comprehensive feature vector. After the LSTM and CNN hybrid model is input into this vector, it outputs the probability of failure in the next hour. The system triggers the early warning linkage module 4 to issue an early warning. According to different early warning levels, the system pushes the early warning and printer operating status information to the equipment administrator via WeChat. The operation and maintenance management platform automatically generates a dispatch task and marks the required tools. At the same time, the system sends optimization instructions to the printer. The early warning data, processing results, and subsequent operation and maintenance records are synchronously stored in the data storage module 5 as samples for model iteration to improve the accuracy of subsequent predictions.

[0028] The following points should be noted in this article: 1. The accompanying drawings of the embodiments of the present invention only involve the structures involved in the embodiments of the present invention; other structures can refer to general designs.

[0029] 2. Where there is no conflict, the embodiments of the present invention and the features thereof can be combined with each other to obtain new embodiments.

[0030] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An artificial intelligence-based printer data monitoring and early warning system, characterized in that, include: The modules include a data acquisition module (1), an edge computing module (2), an AI core analysis module (3), an early warning linkage module (4), and a data storage module (5). Each module interacts with the other via the TCP / IP communication protocol. The data acquisition module (1) is a multi-source data sensing unit. The data acquisition module (1) includes: a working parameter sensor (606), an environmental parameter sensor (605), and an operation behavior acquisition unit. After preprocessing, it outputs a structured data stream. The edge computing module (2) is deployed on the local edge computing node of the printer component (6) and is responsible for data processing, data diversion, real-time threshold judgment and data compression. The AI ​​core analysis module (3) is deployed on a cloud server. The AI ​​core analysis module (3) includes: a feature fusion submodule, a deep learning prediction submodule and a dynamic threshold generation submodule. The early warning linkage module (4) is based on the output results of the AI ​​core analysis module (3) to realize hierarchical early warning and multi-channel linkage, and to link the printer control system, operation and maintenance management platform and consumable management system to perform response operations. The data storage module (5) adopts a hybrid architecture of edge storage and cloud storage, and stores real-time data and historical data in a hierarchical manner.

2. The printer data monitoring and early warning system based on artificial intelligence according to claim 1, characterized in that, The working parameter sensor (606) of the data acquisition module (1) is integrated inside the printer assembly (6) and the API interface of the printer assembly (6). The working parameter sensor (606) collects the operating data of the core components.

3. The printer data monitoring and early warning system based on artificial intelligence according to claim 2, characterized in that, The environmental parameter sensor (605) of the data acquisition module (1) is deployed around the printer assembly (6). The environmental parameter sensor (605) collects ambient temperature and humidity, dust concentration and voltage stability. The operation behavior acquisition unit obtains user operation data through the control chip inside the printer assembly (6). The data acquisition module (1) denoises, standardizes and removes outliers from the collected raw data to generate a structured data stream.

4. The printer data monitoring and early warning system based on artificial intelligence according to claim 3, characterized in that, The edge computing module (2) processes the acquired data, including: data splitting, real-time threshold judgment and data compression; data splitting performs local calculation on high-frequency real-time data and uploads low-frequency statistical data to the cloud AI core analysis module (3); real-time threshold judgment has a built-in lightweight rule engine, which presets basic thresholds based on printer model, and quickly judges sudden extreme value data to trigger immediate warnings; data compression uses the LZ4 compression algorithm to compress the uploaded data.

5. The printer data monitoring and early warning system based on artificial intelligence according to claim 4, characterized in that, The feature fusion submodule of the AI ​​core analysis module (3) uses an attention mechanism to fuse multi-dimensional data features, assigns dynamic weights to hardware parameters, environmental data and operational behavior features, and generates a comprehensive feature vector.

6. The printer data monitoring and early warning system based on artificial intelligence according to claim 5, characterized in that, The deep learning prediction submodule of the AI ​​core analysis module (3) constructs a hybrid model based on LSTM and CNN, where CNN is used to extract spatial features from the data and LSTM is used to capture time series features; the hybrid model of LSTM and CNN is trained and optimized through a large number of fault samples of printer components (6) and common faults, with the input being a comprehensive feature vector and the output being the probability of fault occurrence and fault type in the next hour; The dynamic threshold generation submodule of the AI ​​core analysis module (3) dynamically adjusts the warning thresholds of each parameter based on the printer's historical operating data and real-time operating conditions using a random forest algorithm.

7. The printer data monitoring and early warning system based on artificial intelligence according to claim 6, characterized in that, The warning linkage module (4) classifies the warning into three levels based on the probability of failure and the degree of impact. The warning levels include: Level 1, only prompting through the indicator light on the operation panel of the printer component (6) and simultaneously pushing optimization suggestions; Level 2, potential failure, notifying the equipment administrator via SMS and WeChat, with an attached failure prediction report; Level 3, immediately triggering an audible and visual alarm, automatically pushing dispatch information to the operation and maintenance management platform, and suspending non-emergency printing tasks. The linkage processing of the early warning linkage module (4) is linked with the printer control system. When a first-level early warning occurs, the printing parameters are automatically adjusted. When a third-level early warning occurs, a safety protection operation is performed. The linkage processing is linked with the consumable management system. When a consumable exhaustion fault is predicted, the consumable procurement process is automatically triggered.

8. The printer data monitoring and early warning system based on artificial intelligence according to claim 7, characterized in that, The edge storage of the data storage module (5) stores high-frequency real-time data for the past 7 days locally, using SSD storage to ensure data read and write speed; the cloud storage of the data storage module (5) uses a distributed database to store historical data, providing data support for model iteration and operation and maintenance optimization.

9. The printer data monitoring and early warning system based on artificial intelligence according to claim 8, characterized in that, The printer assembly (6) has an operation panel installed on its top and a storage side groove (601) on its side. A guide rail (602) is installed inside the storage side groove (601). An installation groove (603) is provided on the side end of the guide rail (602), and a positioning slider (604) is slidably installed inside the guide rail (602).

10. The printer data monitoring and early warning system based on artificial intelligence according to claim 9, characterized in that, An environmental parameter sensor (605) is installed at the outer end of the positioning slider (604); the environmental parameter sensor (605) is located inside the receiving side groove (601); a working parameter sensor (606) is installed inside the printer assembly (6).

Images